Carvedilol (Coreg), a beta- and alpha-blocker and an antioxidant drug, was evaluated for moderate to severe heart failure patients in a program containing four United States and one Australia/New Zealand study. The data were evaluated twice by the Cardiovascular and Renal Drugs Advisory Committee of the U.S. Food and Drug Administration (FDA). These meetings resulted in opposite decisions by the advisory committee. The crux of the argumentation was the two-positive-trial FDA paradigm. Carvedilol did not meet the usual paradigm because an exercise end point was not statistically different from placebo in three U.S. trials. Most other end points were highly significant, and death, which was monitored across the U.S. program, was different with p < 0.0001. Here we argue that the usual paradigm is very useful but not an absolute principle, that the usual paradigm can sometimes miss the strength of evidence even in the primary end points, and that rational decision making requires on occasion that other evidence must lead to approval. Control of the type I error rate should be taken very seriously, should rarely be violated, and serves the biomedical community well. It is not an absolute principle, however, but rather must be considered in context.